Parallel Processing with Oracle Scheduler Chains
As i mentioned before, Oracle Scheduler one of the my
favourite utilities of Oracle Database. The capabilities of Scheduler is
limited by your imagination J
The issue that i want to mention now is, how to make a
process to be run in parallel by the help of Chains. Oracle Scheduler Chain is
a set of steps,rules and programs that allows you to design the program blocks.
For instance, you have 4 programs(jobs or processes) the first one is should
run firstly, second and third ones should run after the first one succeeded,
the last one is should run after second and third one is completed(failed or
succeeded) This complex logic can be implemented with Chains easily. It is
really easy, i am not joking. Chains have several APIs and you can use these
API functions as your design needs.
It is time to see Chains in the action…
Business Scenario : We are in the telco sector and we are
implementing something that should be
run before montly billing process. Our telco operator has over 30M subs. So, we
need to implement the something
enough fast…
Implementaion : We have one process that should be run first
in order to find the subscirbers who we are interested in. After the subs have
been found, we can run some business rules over them.
Design : We will create a table(CUTOFF_INTERIM_TABLE) in order
to hold interested-in-subscribers. A simple routine(FillInterimTable procedure) will fill this table. Then we will
use a routine(ProcessInterimTable procedure) to get the subscribers and do something. First procedure(FillInterimTable)
should take an input parameter(pin_TotalProcessCount) and use a simple MODULO
operation(MOD(i,
pin_TotalProcessCount) + 1) to seperate the subscirbers into chunks.
Second procedure(ProcessInterimTable)
can be run (pin_TotalProcessCount)
times simultaneously, and every process will take only their chunks(…WHERE PROCESS_NUMBER =
pin_ProcessNumber…). The important
thing is that, second procedures should run if
and only if the first procedure succeeded. Otherwise there is no need to
run second processes because of there is no subscribers to be processed. We
will use 5 parallel processes.
Tables to be created:
CREATE TABLE
CUTOFF_INTERIM_TABLE
(
CUTOFF DATE,
CO_ID NUMBER,
STATUS VARCHAR2(1),
CREATE_DATE DATE,
REMARK VARCHAR2(1024),
PROCESS_NUMBER NUMBER
);
CREATE TABLE
CUTOFF_PROCESS_EXECUTION_LOG
(
CUTOFF DATE,
PROCESS_NUMBER NUMBER,
LOG_MESSAGE VARCHAR2(1024),
LOG_TIMESTAMP TIMESTAMP
);
Procedures to be created:
CREATE OR REPLACE PROCEDURE
CreateProcessExecutionLog
(
pid_Cutoff IN
CUTOFF_PROCESS_EXECUTION_LOG.CUTOFF%TYPE,
pin_ProcessNumber IN
CUTOFF_PROCESS_EXECUTION_LOG.PROCESS_NUMBER%TYPE,
pis_LogMessage IN CUTOFF_PROCESS_EXECUTION_LOG.LOG_MESSAGE%TYPE
) IS
PRAGMA AUTONOMOUS_TRANSACTION;
BEGIN
INSERT INTO CUTOFF_PROCESS_EXECUTION_LOG
(CUTOFF, PROCESS_NUMBER,
LOG_MESSAGE, LOG_TIMESTAMP)
VALUES
(pid_Cutoff,
pin_ProcessNumber, pis_LogMessage, SYSTIMESTAMP);
COMMIT;
END
CreateProcessExecutionLog;
/
CREATE OR REPLACE PROCEDURE FillInterimTable
(
pid_Cutoff IN CUTOFF_INTERIM_TABLE.CUTOFF%TYPE,
pin_TotalProcessCount IN NUMBER
) IS
vn_ProcessIndependent
CUTOFF_INTERIM_TABLE.PROCESS_NUMBER%TYPE;
vs_LogMessage
CUTOFF_PROCESS_EXECUTION_LOG.LOG_MESSAGE%TYPE;
BEGIN
vn_ProcessIndependent := 0;
vs_LogMessage := 'FillInterimTable started...';
CreateProcessExecutionLog(pid_Cutoff, vn_ProcessIndependent,
vs_LogMessage);
--some business checks etc
---
----
EXECUTE IMMEDIATE 'TRUNCATE TABLE
CUTOFF_INTERIM_TABLE';
FOR i IN 1 .. 100000 LOOP
INSERT INTO CUTOFF_INTERIM_TABLE
(CUTOFF, CO_ID, STATUS,
CREATE_DATE, REMARK, PROCESS_NUMBER )
VALUES
(pid_Cutoff, i, 'N', SYSDATE, 'Successfully
created.', MOD(i,
pin_TotalProcessCount) + 1);
END LOOP;
vs_LogMessage := 'FillInterimTable finished...';
CreateProcessExecutionLog(pid_Cutoff, vn_ProcessIndependent,
vs_LogMessage);
COMMIT;
END
FillInterimTable;
/
CREATE OR REPLACE PROCEDURE
ProcessInterimTable
(
pid_Cutoff IN CUTOFF_INTERIM_TABLE.CUTOFF%TYPE,
pin_ProcessNumber IN CUTOFF_INTERIM_TABLE.PROCESS_NUMBER%TYPE,
pin_TotalProcessCount IN NUMBER
) IS
vs_LogMessage
CUTOFF_PROCESS_EXECUTION_LOG.LOG_MESSAGE%TYPE;
BEGIN
vs_LogMessage := 'ProcessInterimTable started for
process ' ||
pin_ProcessNumber;
CreateProcessExecutionLog(pid_Cutoff, pin_ProcessNumber, vs_LogMessage);
IF pin_TotalProcessCount = 1 AND
pin_ProcessNumber = 1 THEN
-- there is only one process, no
need parallel processing
FillInterimTable(pid_Cutoff,
pin_TotalProcessCount);
END IF;
FOR rec_InterimTable IN (
SELECT T.CO_ID, T.REMARK, T.ROWID AS RD
FROM CUTOFF_INTERIM_TABLE T
WHERE STATUS = 'N'
AND PROCESS_NUMBER =
pin_ProcessNumber
) LOOP
--some business processes
---
----
UPDATE CUTOFF_INTERIM_TABLE
SET STATUS = 'S',
REMARK = 'Successfully
processed..'
WHERE ROWID = rec_InterimTable.RD;
COMMIT;
END LOOP;
vs_LogMessage := 'ProcessInterimTable finished for
process ' ||
pin_ProcessNumber;
CreateProcessExecutionLog(pid_Cutoff, pin_ProcessNumber, vs_LogMessage);
END
ProcessInterimTable;
/
First process Scheduler configuration(5 processes):
DECLARE
vn_TotalProcessCount PLS_INTEGER;
vd_CutOff DATE;
BEGIN
vn_TotalProcessCount := 5;
vd_CutOff := TO_DATE('20.09.2011', 'DD.MM.YYYY');
DBMS_SCHEDULER.create_program(program_name => 'CUTOFF__FILL_PRG',
program_type => 'STORED_PROCEDURE',
program_action => 'FillInterimTable',
number_of_arguments => 2,
enabled => FALSE,
comments => 'Program for
FillInterimTable(5 parallel processes)');
DBMS_SCHEDULER.define_program_argument(program_name => 'CUTOFF__FILL_PRG',
argument_position => 1,
argument_type => 'DATE',
default_value =>
vd_CutOff);
DBMS_SCHEDULER.define_program_argument(program_name => 'CUTOFF__FILL_PRG',
argument_position => 2,
argument_type => 'NUMBER',
default_value =>
vn_TotalProcessCount);
DBMS_SCHEDULER.create_job(job_name
=> 'CUTOFF__FILL_JOB',
program_name => 'CUTOFF__FILL_PRG',
start_date => NULL,
enabled => FALSE,
auto_drop => FALSE,
comments => 'Job for
FillInterimTable(5 parallel processes)');
END;
/
BEGIN
DBMS_SCHEDULER.enable(name => 'CUTOFF__FILL_PRG' );
DBMS_SCHEDULER.enable(name => 'CUTOFF__FILL_JOB' );
END;
/
Second process Scheduler configuration(5 processes):
DECLARE
vn_TotalProcessCount PLS_INTEGER;
vd_CutOff DATE;
BEGIN
vn_TotalProcessCount := 5;
vd_CutOff := TO_DATE('20.09.2011', 'DD.MM.YYYY');
FOR vn_ProcessNumber IN 1..vn_TotalProcessCount
LOOP
DBMS_SCHEDULER.create_program(program_name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_PRG',
program_type => 'STORED_PROCEDURE',
program_action => 'ProcessInterimTable',
number_of_arguments => 3,
enabled => FALSE,
comments => 'Program
ProcessInterimTable ' || vn_ProcessNumber || ' of 5 process');
DBMS_SCHEDULER.define_program_argument(program_name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_PRG',
argument_position => 1,
argument_type => 'DATE',
default_value =>
vd_CutOff);
DBMS_SCHEDULER.define_program_argument(program_name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_PRG',
argument_position => 2,
argument_type => 'NUMBER',
default_value => vn_ProcessNumber);
DBMS_SCHEDULER.define_program_argument(program_name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_PRG',
argument_position => 3,
argument_type => 'NUMBER',
default_value =>
vn_TotalProcessCount);
DBMS_SCHEDULER.create_job(job_name
=> 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_JOB',
program_name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_PRG',
start_date => NULL,
enabled => FALSE,
auto_drop => FALSE,
comments => 'Job for
ProcessInterimTable ' || vn_ProcessNumber || ' of 5 process');
END LOOP;
END;
/
DECLARE
vn_TotalProcessCount PLS_INTEGER;
BEGIN
vn_TotalProcessCount := 5;
FOR vn_ProcessNumber IN 1..vn_TotalProcessCount
LOOP
DBMS_SCHEDULER.enable(name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_JOB');
DBMS_SCHEDULER.enable(name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_PRG');
END LOOP;
END;
/
Scheduler Chain configuration:
BEGIN
DBMS_SCHEDULER.create_chain(chain_name
=> 'CUTOFF__CHAIN',
rule_set_name => NULL,
evaluation_interval => NULL,
comments => 'Chain for
CUTOFF( 5 processes)');
DBMS_SCHEDULER.define_chain_step(chain_name => 'CUTOFF__CHAIN',
step_name => 'FILL_INTERIM_TABLE_STEP',
program_name => 'CUTOFF__FILL_PRG');
DBMS_SCHEDULER.define_chain_step(chain_name => 'CUTOFF__CHAIN',
step_name => 'EXECUTE_1OF5_STEP',
program_name => 'CUTOFF__EXEC1_PRG');
DBMS_SCHEDULER.define_chain_step(chain_name => 'CUTOFF__CHAIN',
step_name => 'EXECUTE_2OF5_STEP',
program_name => 'CUTOFF__EXEC2_PRG');
DBMS_SCHEDULER.define_chain_step(chain_name => 'CUTOFF__CHAIN',
step_name => 'EXECUTE_3OF5_STEP',
program_name => 'CUTOFF__EXEC3_PRG');
DBMS_SCHEDULER.define_chain_step(chain_name => 'CUTOFF__CHAIN',
step_name => 'EXECUTE_4OF5_STEP',
program_name => 'CUTOFF__EXEC4_PRG');
DBMS_SCHEDULER.define_chain_step(chain_name => 'CUTOFF__CHAIN',
step_name => 'EXECUTE_5OF5_STEP',
program_name => 'CUTOFF__EXEC5_PRG');
END;
/
BEGIN
DBMS_SCHEDULER.define_chain_rule(chain_name => 'CUTOFF__CHAIN',
condition => 'TRUE',
action => 'START
FILL_INTERIM_TABLE_STEP',
rule_name => 'FILL_INTERIM_TABLE_RULE',
comments => 'Start CUTOFF -
Fill Interim Table.');
DBMS_SCHEDULER.define_chain_rule(chain_name => 'CUTOFF__CHAIN',
condition => 'FILL_INTERIM_TABLE_STEP
SUCCEEDED',
action => 'START
EXECUTE_1OF5_STEP',
rule_name => 'EXECUTE_1OF5_RULE',
comments => 'Execute CUTOFF -
1 of 5 process.');
DBMS_SCHEDULER.define_chain_rule(chain_name => 'CUTOFF__CHAIN',
condition => 'FILL_INTERIM_TABLE_STEP
SUCCEEDED',
action => 'START
EXECUTE_2OF5_STEP',
rule_name => 'EXECUTE_2OF5_RULE',
comments => 'Execute CUTOFF -
2 of 5 process.');
DBMS_SCHEDULER.define_chain_rule(chain_name => 'CUTOFF__CHAIN',
condition => 'FILL_INTERIM_TABLE_STEP
SUCCEEDED',
action => 'START
EXECUTE_3OF5_STEP',
rule_name => 'EXECUTE_3OF5_RULE',
comments => 'Execute CUTOFF -
3 of 5 process.');
DBMS_SCHEDULER.define_chain_rule(chain_name => 'CUTOFF__CHAIN',
condition => 'FILL_INTERIM_TABLE_STEP
SUCCEEDED',
action => 'START
EXECUTE_4OF5_STEP',
rule_name => 'EXECUTE_4OF5_RULE',
comments => 'Execute CUTOFF -
4 of 5 process.');
DBMS_SCHEDULER.define_chain_rule(chain_name => 'CUTOFF__CHAIN',
condition => 'FILL_INTERIM_TABLE_STEP
SUCCEEDED',
action => 'START
EXECUTE_5OF5_STEP',
rule_name => 'EXECUTE_5OF5_RULE',
comments => 'Execute CUTOFF -
5 of 5 process.');
DBMS_SCHEDULER.define_chain_rule(chain_name => 'CUTOFF__CHAIN',
condition => '(FILL_INTERIM_TABLE_STEP
FAILED) OR (EXECUTE_1OF5_STEP COMPLETED AND EXECUTE_2OF5_STEP COMPLETED AND
EXECUTE_3OF5_STEP COMPLETED AND EXECUTE_4OF5_STEP COMPLETED AND
EXECUTE_5OF5_STEP COMPLETED)',
action => 'END',
rule_name => 'EXECUTE_END_RULE',
comments => 'Execute CUTOFF -
finished.');
END;
/
BEGIN
DBMS_SCHEDULER.ENABLE('CUTOFF__CHAIN');
END;
/
Job configuraion that will be used to invoke Chain:
BEGIN
DBMS_SCHEDULER.create_job(job_name
=> 'CUTOFF__CHAIN_JOB',
job_type => 'CHAIN',
job_action => 'CUTOFF__CHAIN',
start_date => NULL,
enabled => FALSE,
auto_drop => FALSE,
comments => 'Job for
CUTOFF__CHAIN ');
END;
/
Every thing is OK.
We can test our implementation.
TRUNCATE TABLE
CUTOFF_INTERIM_TABLE;
DELETE FROM
CUTOFF_PROCESS_EXECUTION_LOG;
COMMIT;
BEGIN
DBMS_SCHEDULER.run_job(job_name
=> 'CUTOFF__CHAIN_JOB',use_current_session
=> FALSE);
END;
/
Let’s check running chain:
SQL> SELECT step_name, state FROM USER_SCHEDULER_RUNNING_CHAINS WHERE
CHAIN_NAME = 'CUTOFF__CHAIN' order by start_date;
STEP_NAME
STATE
------------------------------ ---------------
FILL_INTERIM_TABLE_STEP
RUNNING
EXECUTE_2OF5_STEP
NOT_STARTED
EXECUTE_3OF5_STEP
NOT_STARTED
EXECUTE_4OF5_STEP
NOT_STARTED
EXECUTE_5OF5_STEP
NOT_STARTED
EXECUTE_1OF5_STEP
NOT_STARTED
6 rows selected
SQL> SELECT step_name, state FROM USER_SCHEDULER_RUNNING_CHAINS WHERE
CHAIN_NAME = 'CUTOFF__CHAIN' order by start_date;
STEP_NAME
STATE
------------------------------ ---------------
FILL_INTERIM_TABLE_STEP
SUCCEEDED
EXECUTE_2OF5_STEP
RUNNING
EXECUTE_3OF5_STEP
RUNNING
EXECUTE_4OF5_STEP
NOT_STARTED
EXECUTE_5OF5_STEP
NOT_STARTED
EXECUTE_1OF5_STEP
NOT_STARTED
6 rows selected
SQL> SELECT step_name, state FROM USER_SCHEDULER_RUNNING_CHAINS WHERE
CHAIN_NAME = 'CUTOFF__CHAIN' order by start_date;
STEP_NAME
STATE
------------------------------ ---------------
FILL_INTERIM_TABLE_STEP SUCCEEDED
EXECUTE_2OF5_STEP
RUNNING
EXECUTE_3OF5_STEP
RUNNING
EXECUTE_1OF5_STEP
RUNNING
EXECUTE_4OF5_STEP
RUNNING
EXECUTE_5OF5_STEP
RUNNING
6 rows selected
SQL> SELECT step_name, state FROM USER_SCHEDULER_RUNNING_CHAINS WHERE
CHAIN_NAME = 'CUTOFF__CHAIN' order by start_date;
STEP_NAME
STATE
------------------------------ ---------------
FILL_INTERIM_TABLE_STEP
SUCCEEDED
EXECUTE_2OF5_STEP
RUNNING
EXECUTE_3OF5_STEP
RUNNING
EXECUTE_1OF5_STEP
RUNNING
EXECUTE_4OF5_STEP
RUNNING
EXECUTE_5OF5_STEP
RUNNING
6 rows selected
SQL> SELECT step_name, state FROM USER_SCHEDULER_RUNNING_CHAINS WHERE
CHAIN_NAME = 'CUTOFF__CHAIN' order by start_date;
STEP_NAME
STATE
------------------------------ ---------------
FILL_INTERIM_TABLE_STEP
SUCCEEDED
EXECUTE_2OF5_STEP
SUCCEEDED
EXECUTE_3OF5_STEP SUCCEEDED
EXECUTE_1OF5_STEP
SUCCEEDED
EXECUTE_4OF5_STEP
RUNNING
EXECUTE_5OF5_STEP
SUCCEEDED
6 rows selected
SQL> SELECT step_name, state FROM USER_SCHEDULER_RUNNING_CHAINS WHERE
CHAIN_NAME = 'CUTOFF__CHAIN' order by start_date;
STEP_NAME
STATE
------------------------------ ---------------
SQL>
SQL> SELECT job_subname, status,actual_start_date, additional_info
FROM USER_SCHEDULER_JOB_RUN_DETAILS WHERE JOB_NAME = 'CUTOFF__CHAIN_JOB'
ORDER BY log_id DESC;
JOB_SUBNAME STATUS ACTUAL_START_DATE
ADDITIONAL_INFO
---------------------------
--------------- -------------------------------------------------
----------------------------------------------------------
SUCCEEDED 30/10/2011 16:22:32,068405 +02:00
CHAIN_LOG_ID="333627"
EXECUTE_4OF5_STEP SUCCEEDED 30/10/2011 16:22:37,258659
AFRICA/CAIRO
CHAIN_LOG_ID="333627", STEP_NAME="EXECUTE_4OF5_STEP"
EXECUTE_2OF5_STEP SUCCEEDED 30/10/2011 16:22:37,198698
AFRICA/CAIRO
CHAIN_LOG_ID="333627", STEP_NAME="EXECUTE_2OF5_STEP"
EXECUTE_5OF5_STEP SUCCEEDED 30/10/2011 16:22:37,318665
AFRICA/CAIRO
CHAIN_LOG_ID="333627", STEP_NAME="EXECUTE_5OF5_STEP"
EXECUTE_1OF5_STEP SUCCEEDED 30/10/2011 16:22:37,108348
AFRICA/CAIRO
CHAIN_LOG_ID="333627", STEP_NAME="EXECUTE_1OF5_STEP"
EXECUTE_3OF5_STEP SUCCEEDED 30/10/2011 16:22:37,206311
AFRICA/CAIRO CHAIN_LOG_ID="333627",
STEP_NAME="EXECUTE_3OF5_STEP"
FILL_INTERIM_TABLE_STEP SUCCEEDED 30/10/2011 16:22:32,112917
AFRICA/CAIRO
CHAIN_LOG_ID="333627", STEP_NAME="FILL_INTERIM_TABLE_STEP"
SQL> SELECT PROCESS_NUMBER, LOG_MESSAGE, LOG_TIMESTAMP FROM
CUTOFF_PROCESS_EXECUTION_LOG ORDER BY LOG_TIMESTAMP;
PROCESS_NUMBER
LOG_MESSAGE
LOG_TIMESTAMP
--------------
-------------------------------------------- ----------------------------
0 FillInterimTable started... 30/10/2011 16:22:32,210113
0 FillInterimTable
finished... 30/10/2011
16:22:37,076799
2 ProcessInterimTable started for
process 2 30/10/2011 16:22:37,200165
3 ProcessInterimTable started for
process 3 30/10/2011 16:22:37,229709
1 ProcessInterimTable started for
process 1 30/10/2011 16:22:37,338941
4 ProcessInterimTable started for
process 4 30/10/2011 16:22:37,530999
5 ProcessInterimTable started for
process 5 30/10/2011 16:22:37,631659
3 ProcessInterimTable finished for
process 3 30/10/2011 16:22:47,462734
1 ProcessInterimTable finished for
process 1 30/10/2011 16:22:48,243164
5 ProcessInterimTable finished for
process 5 30/10/2011 16:22:48,411946
2 ProcessInterimTable finished for
process 2 30/10/2011 16:22:48,470194
4 ProcessInterimTable finished for
process 4 30/10/2011 16:22:48,489575
12 rows selected
SQL>
As you can see above, a few configuration is enough
in order to implement our design.
Please note that i have used cutoff parameter
hard-coded in the programs above. This is because of restrictions of Oracle
Chains. It is not possible to set job argument values on the fly with chains. I
have made some tricks to avoid this situation. I will share this trick later J
To rollback everything:
DROP TABLE
CUTOFF_PROCESS_EXECUTION_LOG;
DROP TABLE
CUTOFF_INTERIM_TABLE;
DROP PROCEDURE
CreateProcessExecutionLog;
DROP PROCEDURE
FillInterimTable;
DROP PROCEDURE
ProcessInterimTable;
BEGIN
DBMS_SCHEDULER.drop_chain(chain_name
=> 'CUTOFF__CHAIN',force => true);
END;
/
BEGIN
DBMS_SCHEDULER.purge_log(job_name => 'CUTOFF__CHAIN_JOB');
DBMS_SCHEDULER.drop_job(job_name
=> 'CUTOFF__CHAIN_JOB');
END;
/
BEGIN
DBMS_SCHEDULER.purge_log(job_name => 'CUTOFF__FILL_JOB');
DBMS_SCHEDULER.drop_job(job_name
=> 'CUTOFF__FILL_JOB');
DBMS_SCHEDULER.drop_program(program_name => 'CUTOFF__FILL_PRG');
END;
/
DECLARE
vn_TotalProcessCount PLS_INTEGER;
BEGIN
vn_TotalProcessCount := 5;
FOR vn_ProcessNumber IN 1..vn_TotalProcessCount
LOOP
DBMS_SCHEDULER.purge_log(job_name
=> 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_JOB');
DBMS_SCHEDULER.drop_job(job_name
=> 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_JOB');
DBMS_SCHEDULER.drop_program(program_name => 'CUTOFF__EXEC' ||
vn_ProcessNumber || '_PRG');
END LOOP;
END;
/
Useful queries to check Scheduler objects:
SELECT * FROM
USER_SCHEDULER_JOBS WHERE
JOB_NAME LIKE 'CUTOFF%';
SELECT * FROM
USER_SCHEDULER_JOB_ARGS WHERE JOB_NAME LIKE 'CUTOFF%';
SELECT * FROM
USER_SCHEDULER_PROGRAMS WHERE PROGRAM_NAME LIKE 'CUTOFF%';
SELECT * FROM
USER_SCHEDULER_PROGRAM_ARGS WHERE PROGRAM_NAME LIKE 'CUTOFF%';
SELECT * FROM USER_SCHEDULER_CHAINS
WHERE CHAIN_NAME LIKE 'CUTOFF%';
SELECT * FROM USER_SCHEDULER_CHAIN_STEPS
WHERE CHAIN_NAME LIKE 'CUTOFF%';
SELECT * FROM USER_SCHEDULER_CHAIN_RULES
WHERE CHAIN_NAME LIKE 'CUTOFF%';
SELECT * FROM
USER_SCHEDULER_RUNNING_JOBS WHERE JOB_NAME LIKE 'CUTOFF%';
SELECT * FROM
USER_SCHEDULER_RUNNING_CHAINS WHERE CHAIN_NAME = 'CUTOFF__CHAIN' order by start_date;
SELECT * FROM
USER_SCHEDULER_JOB_LOG WHERE JOB_NAME LIKE 'CUTOFF%' ORDER BY 1 DESC;
SELECT * FROM
USER_SCHEDULER_JOB_RUN_DETAILS WHERE JOB_NAME LIKE 'CUTOFF%' ORDER BY log_id DESC;
SELECT * FROM
SYS.SCHEDULER$_EVENT_LOG /*WHERE LOG_ID = 333551*/;
2 comments:
AWESOME! GREAT EXAMPLE !
- Can we use Procedures in a Package to be run in a program?
- Why create a job for a program if we are using program names in chain?
Post a Comment